JP2013015990A - Information processing system, information processing program and information processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce processing loads required for deleting duplex data in a system duplexing data.SOLUTION: A first processor 11 sequentially stores data in a first memory 14 and transmits storage object data in the first memory 14 to a second processor 21. The first processor 11 sequentially reads the data stored in the first memory 14 from the first memory 14 and stores the data in a hard disk 15. The second processor 21 stores the storage object data transmitted from the first processor 11 in a second memory 24. At this stage, the first processor 11 transmits identification information capable of identifying the data stored lastly in second storage means to the second processor 21. The second processor 21 deletes data identified by the identification information and data stored earlier than the data, of the data stored in the second memory 24.

Description

  The present invention relates to a technique for duplicating data stored in a storage unit of an information processing apparatus.

  In an information processing system, data stored in a storage unit included in a certain node is duplicated in a storage unit included in another node to duplicate the data. As an example of a technique for performing such data duplication, a technique relating to a remote copy system including a primary storage system and a secondary storage system is disclosed. In this technique, when the primary storage system receives a write request, the write data is given a write time and transferred to the secondary storage system. Then, the secondary storage system stores the write data in the logical volume in the secondary storage system based on the write time.

JP 2006-221487 A

  Here, as a case where data is duplexed, a case where data is duplexed in a so-called write-back process is considered as follows. For example, when the first node stores data in the storage means, the data is temporarily stored in the first storage means (for example, storage means capable of performing storage processing at a relatively high speed). deep. Then, the first node sequentially stores data from the first storage unit to the second storage unit (for example, a storage unit that requires a relatively long storage process). At this time, in preparation for a situation in which data is lost from the first storage means before storing the data in the second storage means, the first node is transferred to the third storage means of another second node. The data in the first storage means is duplicated. In this case, when the first node stores the data from the first storage unit to the second storage unit, the stored data is deleted from the first storage unit, and the third node A request is sent to delete the corresponding data duplicated in the storage means. Then, in the second node, the corresponding data is deleted from the third storage means.

  An object of one or more embodiments of the present technology is to reduce a processing load required to delete duplicated data in a system that duplicates such data.

  In the information processing system including the first node and the second node, the present technology performs the following processing. That is, the first node sequentially stores data in the first storage unit and transmits data to be stored in the first storage unit to the second node. In addition, the first node sequentially reads data stored in the first storage unit from the first storage unit and stores the data in the second storage unit. On the other hand, the second node stores the storage target data transmitted from the first node in the third storage means. Here, the first node transmits identification information that can identify the specific data stored in the second storage means to the second node. Then, the second node deletes the specific data identified by the identification information from the data stored in the third storage unit and the data stored before the specific data.

  According to the present technology, in one aspect, in a system that performs data duplication, it is possible to reduce the processing load required to delete the duplicated data.

It is explanatory drawing which shows an example of the functional block of an information processing system. It is explanatory drawing of an example of the data structure of a 1st memory, a 2nd memory, and a duplication request | requirement. It is explanatory drawing of an example of the sequence of a duplication process. It is explanatory drawing of the specific example of the data in a duplication process. It is a flowchart which shows an example of the process in the 1st processor side (at the time of write request reception). It is explanatory drawing of an example of the allocation process part of a 1st memory. It is explanatory drawing of an example of the allocation process part of a 1st memory. It is a flowchart which shows an example of the process (data storage in a hard disk) in the 1st processor side. It is a flowchart which shows an example of the process in the 2nd processor side (at the time of write request reception). It is explanatory drawing of an example of a hardware configuration.

[1. Outline of Embodiment]
In the embodiment described in this specification, when a write-back process is performed in a certain processor, the processing load required for deleting the duplicated data is reduced as a whole in a system that duplicates dirty data with the storage means of another processor. Will be described.

  Normally, in write-back processing, for example, when a processor of a computer (an example of a first node, hereinafter referred to as a first processor) receives a data write request to a hard disk or the like from another host computer. The following processing is performed. That is, first, data is temporarily stored in an internal storage device such as a memory (which is an example of a first storage unit, and hereinafter referred to as a first memory) that can perform a writing process quickly. Thereafter, the first processor stores the data stored in the first memory in an external storage device (an example of the second storage means) such as a hard disk, and deletes the corresponding data in the first memory. . Data that has been stored in the memory or the like but not yet stored in the hard disk or the like is called dirty data. Here, before storing data in the hard disk, the first processor returns a response indicating completion of writing at the time when the data is stored in the first memory. This is because a response to the host computer can be quickly performed.

  However, the data temporarily stored in the first memory may be lost for some reason after the first processor temporarily stores the data in the first memory until it is stored in the hard disk. In this case, although the write completion response has already been transmitted to the host computer, the data is actually lost without being stored in the hard disk. In order to avoid such a situation, when the first processor stores the data in the first memory, this first processor is compared with other processors (an example of the second node, hereinafter referred to as the second processor). Send a data duplication request. Then, on the second processor side that has received the dirty data, this data is stored in a second memory (an example of second storage means, hereinafter referred to as a second memory). In this way, the data is duplicated in the first memory and the second memory, and when the data in the first memory on the first processor side is lost, it is stored in the second memory on the second processor side. Use data to store to hard disk.

  Here, in such a system in which dirty data is duplicated, when the first processor stores the data in the hard disk and the corresponding data in the first memory is deleted, the data is naturally stored in the second memory on the second processor side. The necessary data is no longer necessary. Therefore, when the first processor stores the data in the hard disk and deletes the corresponding data in the first memory, the first processor transmits a corresponding data deletion request also to the second processor. When the second processor receives this deletion request, the corresponding data stored in the second memory is deleted (purged). That is, the first processor transmits a deletion request to the second processor every time data in the first memory is stored in the hard disk. On the other hand, every time the second processor receives the deletion request, the second processor deletes the data in the second memory.

  The technique described in the embodiment of the present specification eliminates the need to send a deletion request to the second processor every time data in the first memory is stored in the hard disk in this way. It is a technology that makes it possible to delete a plurality of data at once, and reduces the processing load required for deleting duplicated data as a whole system.

  Specifically, in the present technology, the first processor transmits identification information that can identify data last stored in the hard disk to the second processor. In other words, the “identification information that can identify the data stored last in the hard disk” identifies the data stored most recently among the data stored in the first memory of the first processor. It is also possible information. That is, if the most recently stored data among the data stored in the first memory of the first processor can be identified, the previous data is the data stored last in the hard disk. . On the other hand, the second processor deletes the data stored before this data, including the data stored in the hard disk last identified by the identification information, from the second memory.

  More specifically, in the present technology, when data is stored in the first memory in the first processor, storage order information indicating the order in which the data is stored in the first memory is associated. On the other hand, when the first processor transmits the data duplication request to the second processor, the first processor adds the storage order information associated with the data and transmits it. Furthermore, the first processor adds the storage order information associated with the data stored most recently among the data stored in the first memory to the data, and transmits it as identification information. In the second processor, the received data is stored in the second memory together with the storage order information associated with the data, while the identification received from the first processor among the data stored in the second memory. Based on the information, the data stored before this data, including the data stored last on the hard disk, is deleted from the second memory.

  As described above, in the present technology, since the first processor transmits the identification information to the second processor, data to be deleted from the second memory can be identified and deleted on the second processor side. At this time, the second processor can delete data stored before this data from the second memory, including the data stored last in the hard disk, based on this identification information. It is possible to delete. Further, even if there is only one data that can be deleted in a single measure (that is, only the data stored last in the hard disk), at least every time the first processor stores the data in the first memory in the hard disk. There is no need to send a delete request to the second processor. For this reason, it is possible to reduce the processing load of the entire system.

[2. Description of overall configuration]
FIG. 1 is a functional block diagram of an example of an information processing system according to the present embodiment.
The information processing system includes a first computer 1, a second computer 2, and a host computer 3. The first computer 1 and the second computer 2 and the first computer 1 and the host computer 3 are network-connected so as to be communicable by, for example, a LAN (Local Area Network). The first computer 1 includes a first processor 11, and the second computer 2 includes a second processor 21. Further, the first computer 1 includes a hard disk 15 which is an example of a storage unit.

The first processor 11 includes a first memory storage control unit 12, a disk storage control unit 13, and a first memory 14.
The first memory storage control unit 12 temporarily stores the data included in the write request received from the host computer 3 in the first memory 14 in association with the storage time that is the current time. At this time, the first memory storage control unit 12 transmits a data duplication request of the first memory 14 to the second processor 21.

  Specifically, when receiving a data write request to the hard disk 15 from the host computer 3, the first memory storage control unit 12 sets the current time as the storage time and sets the storage time to the data included in the write request. The added duplex request is transmitted to the second processor 21 of the second computer 2. This storage time is an example of storage order information indicating the storage order of data. At this time, the first memory storage control unit 12 adds the storage time associated with the most recently stored data among the data existing in the first memory 14 to the duplex request as the oldest time. 2 is transmitted to the second processor 21 of the computer 2. This oldest time is an example of identification information that can identify the data stored in the hard disk 15 last. In other words, it can be said that the first memory storage control unit 12 transmits the oldest time in synchronization with the duplex request. The first memory storage control unit 12 stores the data included in the write request from the host computer 3 in the first memory 14 when receiving the response that the duplication is completed from the second processor 21. At this time, the first memory storage control unit 12 stores the above-mentioned storage time in association with the data included in the write request.

The disk storage control unit 13 sequentially reads the data stored in the first memory 14 in order of oldest storage time and stores the data in the hard disk 15 at predetermined time intervals.
The second processor 21 includes a second memory storage control unit 22, a deletion unit 23, and a second memory 24.

  When the second memory storage control unit 22 receives the duplex request from the first processor 11, among the data included in the received duplex request, the data included in the write request and the storage time added to the data are Store in the second memory 24.

  When receiving the duplex request from the first processor 11, the deletion unit 23 deletes data stored in the second memory 24 and having a storage time earlier than the oldest time included in the duplex request.

[3. First memory and second memory, duplex request data structure]
Next, the data stored in the first memory 14 and the second memory 24 and the data structure of the duplex request transmitted from the first processor 11 to the second processor 21 will be described in detail with reference to FIG. explain.

  FIG. 2A shows an example of the data structure of data stored in the first memory 14. The data stored in the first memory 14 includes storage time, device ID, LBA (Local Block Addressing), size, and data contents. As described above, the storage time indicates the time when the data is stored in the first memory 14. The device ID, LBA, size, and data body are data included in the write request received from the host computer 3. The device ID is an identifier that identifies a device that stores data. In this embodiment, the device ID is an identifier indicating the hard disk 15. The LBA is an address indicating a data storage location in the hard disk 15. The size indicates the data size of the data body part. The data body is a body portion (body) of specific data to be stored in the hard disk 15.

  FIG. 2B shows a data structure of data included in the duplex request transmitted from the first processor 11 to the second processor 21. The duplex request data includes the oldest time, storage time, device ID, LBA, size, and data contents. As described above, the earliest time is the storage time of the most recently stored data among the data stored in the first memory 14 when the data is transmitted to the second processor 21. The storage time, device ID, LBA, size, and data content are the same as the data stored in the first memory 14 shown in FIG.

  The structure of data stored in the second memory 24 is the same as the structure of data stored in the first memory 14 shown in FIG. That is, the second memory storage control unit 22 of the second processor 21 receives the data of the other part except the oldest time from the data having the structure shown in FIG. 2B received from the first processor 11. And stored in the second memory 24.

[4. Overview of duplex processing sequence]
Next, referring to the sequence diagram of FIG. 3 for an example of the processing flow when the data stored in the first memory 14 by the first processor 11 of the first computer 1 is duplicated in the second memory 21 of the second processor While explaining.

  The host computer 3 transmits a data write request to the first processor 11 of the first computer 1. This write request includes the device ID, LBA, size, and data body.

  The first memory storage control unit 12 of the first processor 11 transmits a duplex request in which the storage time and the oldest time are added to the data included in the write request to the second processor of the second computer 2.

  When receiving the duplex request, the deleting unit 23 of the second processor 21 deletes data stored in the second memory 24 and having a storage time before the received oldest time. When the second memory storage control unit 22 receives the duplex request, the second memory 24 stores the device ID, the LBA, the size, the data body, and the storage time added to the data among the data included in the duplex request. To store. Then, a response indicating completion of writing to the second memory 24 is returned to the first processor 11.

  When receiving the response from the second processor 21, the first memory storage control unit 12 of the first processor 11 adds the storage time to the device ID, LBA, size, and data body, and stores them in the first memory 14.

  The first processor 11 of the first computer 1 that has received the response from the second computer 2 further returns a response indicating that the writing of the target data of the write request has been completed to the host computer 3.

[5. Data example of duplex processing]
Here, the above-described duplex processing will be further described with reference to a specific example of data in FIG.
First, a state in which a write request is received from the host computer 3 and the first memory storage control unit 12 of the first processor 11 stores data in the first memory 14 is a state illustrated in FIG. The data at the storage time “2011/1/1 12:05:00” is the data that was received and stored. The storage time of the data stored in the first memory 14 with the oldest storage time is “2011/1/1 12:02:00”. In this case, “2011/1/1 12:02:00” is the oldest time.

  At this time, the duplex request transmitted from the first memory storage controller 12 of the first processor 11 to the second processor 21 is as shown in FIG. That is, in the duplex request, the oldest time “2011/1/1 12:02:00” is added to the data at the storage time “2011/1/1 12:05:00” shown in FIG. It becomes the added data.

  Furthermore, it is assumed that the second memory 24 in a state where the duplex request is received and the duplex request data is stored in the second memory 24 is in the state shown in FIG. Here, referring to the data stored in the second memory 24, there are two pieces of data whose storage time is earlier than the oldest time “2011/1/1 12:02:00”. For this reason, the deletion unit 23 of the second processor 21 deletes these two pieces of data and releases the memory area.

  In addition, the processing of the deletion unit 23 is as follows. That is, by including the earliest time “2011/1/1 12:02:00” in the duplex request, the deletion unit 23 stores the oldest time among the data stored in the second memory 24. It is possible to identify that the data at the storage time “2011/1/1 12:01:00” that is earlier than the time and the latest is the data stored in the hard disk 15 last. For this reason, the deletion unit 23 deletes the data identified as described above at the storage time “2011/1/1 12:01:00” and the data with the storage time earlier than this, and releases the memory area.

[6. Detailed description of processing of first processor and second processor]
Next, specific processing executed by each of the first processor 11 and the second processor 21 will be described with reference to FIGS.

<Processing on the first processor side (when receiving a write request)>
FIG. 5 shows processing of the first processor 11 when a data write request is received from the host computer 3.

  In S <b> 1, the first memory storage control unit 12 waits for reception of a data write request from the host computer 3.

  In S <b> 2, the first memory storage control unit 12 receives a data write request from the host computer 3.

  In S3, the first memory storage control unit 12 adds the storage time and the oldest time to the device ID, LBA, size, and data body received from the host computer 3, and transmits them to the second processor 21 as a duplex request. Then, the first memory storage control unit 12 waits for a response from the second processor 21.

  In S4, the first memory storage control unit 12 determines whether or not the response from the second processor 21 is an OK response in which duplication has been normally completed. If it is an OK response (Yes), the process proceeds to S6, and if it is an NG response for which duplication has failed (No), the process proceeds to S5.

  In S <b> 5, the disk storage control unit 13 reads out the data with the oldest storage time from the data stored in the first memory 14 and stores it in the hard disk 15. Then, the first memory storage control unit 12 deletes the data stored in the hard disk 15 from the first memory 14. In other words, the first memory storage control unit 12 releases the area of the first memory 14 in which the data is stored. Then, the process of S3 is executed again.

  In S6, based on the device ID, LBA, and size included in the write request received from the host computer 3, the first memory storage control unit 12 stores the data area to be written included in the write request. It is checked whether data is stored on the first memory 14 or not.

  In S <b> 7, the first memory storage control unit 12 determines whether the data of the disk area in which the data body to be written is stored is stored in the first memory 14 as the check result in S <b> 6. If stored (Yes), the process proceeds to S8. If not stored (No), the process proceeds to S9.

  In S8, the first memory storage control unit 12 determines whether or not a new memory area needs to be allocated when the data included in the write request is stored in the first memory 14. If it is not necessary to allocate a new memory area (Yes), the process proceeds to S9. If a new memory area needs to be allocated (No), the process proceeds to S11.

  The reason why the processes of S6 to S8 are performed is as follows. That is, if data to be stored in the same disk area of the hard disk 15 already exists on the first memory 14, the data body of the data on the first memory 14 is overwritten with the data body of the newly stored data. This is because it is not necessary to separately allocate the memory area of the first memory 14. In addition, the number of times data is stored in the hard disk 15 can be reduced by overwriting the data body of the new data over the existing data and integrating the data.

  In S9, the first memory storage control unit 12 attempts to allocate a memory area in the first memory 14 for storing data included in the write request.

  In S10, the first memory storage control unit 12 determines whether or not the first memory 14 has successfully allocated a memory area for storing data included in the write request. If the allocation is successful (Yes), the process proceeds to S12, and if the allocation fails (No), the process proceeds to S11.

  In S <b> 11, the disk storage control unit 13 reads out the data with the oldest storage time from the data stored in the first memory 14 and stores it in the hard disk 15. Then, the first memory storage control unit 12 releases the area of the first memory 14 in which the data is stored. Then, the process of S9 is executed again.

  In S12, the first memory storage control unit 12 associates the device ID, LBA, size, and data body included in the write request with the storage time of S3 and stores them in an appropriate memory area of the first memory 14. Here, storing in an appropriate memory area means storing in a new memory area when a new memory area is allocated, and if allocation of a new memory area is unnecessary, Indicates that the data body included in the write request is overwritten on the data body of the existing data.

  In S13, the first memory storage controller 12 overwrites the data included in the write request in the memory area of the existing data in S12, and overwrites the entire data body of the existing data in the first memory 14 When this happens, the storage time of the existing data is overwritten with the storage time of S3. Note that this overwriting method is merely an example. For example, the storage time may be overwritten when a part of the data body is overwritten.

  In S <b> 14, the first memory storage control unit 12 returns an OK response indicating completion of writing to the host computer 3.

In addition, the allocation process of the memory area of the first memory 14 in the above-described S6 to S8 or S34 to 36 will be described using a specific example with reference to FIGS.
For example, it is assumed that the data stored in the first memory 14 is the data shown in FIG. This data is device ID = “DEV-A”, LBA = “100”, and size = “100”.

  The data included in the new write request is, for example, as shown in FIG. 6B, device ID = “DEV-A”, LBA = “100”, and size = “100”. The disk area in which the data body of this data is to be stored is the same as the existing data in FIG. Therefore, it is not necessary to allocate a new area in the first memory 14, and the first memory storage control unit 12 overwrites the data body of the existing data with the data body that is the target of the write request. As a result, the data in the first memory 14 is the data shown in FIG. At this time, the first memory storage control unit 12 overwrites the entire data body of the existing data, so the storage time is overwritten with the new storage time (S13).

  Next, when the data included in the new write request is, for example, as shown in FIG. 6C, device ID = “DEV-A”, LBA = “125”, and size = “50”. To do. As shown in FIG. 6C, the disk area in which the data body of this data is stored partially overlaps with the existing data in FIG. 6A, and is stored in a new area that does not exist in the existing data. There is no data to do. Therefore, it is not necessary to allocate a new area in the first memory 14, and the first memory storage control unit 12 overwrites a part of the data body of the existing data with the data body that is the target of the write request. . As a result, the data in the first memory 14 is the data shown in FIG. At this time, since the first memory storage control unit 12 overwrites only a part of the data body of the existing data, the storage time remains the existing data (S13).

  Next, assume that the data included in the new write request is, for example, as shown in FIG. 7A, device ID = “DEV-A”, LBA = “50”, and size = “200”. The disk area in which the data body of this data is to be stored partially overlaps with the existing data in FIG. 6A, and a disk area that does not exist in the existing data is also targeted for writing. Therefore, the first memory storage control unit 12 allocates a new memory area in the first memory 14 and stores data separately.

  Next, it is assumed that the target data of the new write request is, for example, as shown in FIG. 6B, device ID = “DEV-A”, LBA = “125”, and size = “100”. The disk area in which the data body of this data is to be stored partially overlaps with the existing data in FIG. 6A, and a disk area that does not exist in the existing data is also targeted for writing. Therefore, the first memory storage control unit 12 allocates a new memory area in the first memory 14 and stores data separately.

<Processing on the first processor side (data storage in hard disk)>
FIG. 8 shows a data storage process to the hard disk 15 executed by the first processor 11 every predetermined time. The predetermined time is a time set in advance by a system administrator or the like, for example.

  In S <b> 21, the disk storage control unit 13 determines whether data that has not been stored in the hard disk 15 for a predetermined time or longer exists in the first memory 14. Note that this determination can be made, for example, by comparing the time obtained by subtracting a predetermined time from the current time with the data storage time of the first memory 14. If there is data that has not been stored in the hard disk 15 for a predetermined time or longer (Yes), the process proceeds to S22, and if it does not exist, the process ends.

  In S <b> 22, the disk storage control unit 13 reads out the data with the oldest storage time from the data stored in the first memory 14 and stores it in the hard disk 15. Then, the first memory storage control unit 12 releases the area of the first memory 14 in which the data is stored.

<Processing on the second processor side (when receiving a write request)>
FIG. 9 shows processing of the second processor 21 when a data write request is received from the host computer 3.

  In S <b> 31, the second memory storage control unit 22 waits for reception of a duplex request from the first processor 11 of the first computer 1.

  In S <b> 32, the second memory storage control unit 22 receives the duplex request from the first processor 11.

  In S <b> 33, the second memory storage control unit 22 refers to the oldest time included in the duplex request received from the first processor 11. Then, the second memory storage control unit 22 deletes data stored in the second memory 24 and having a storage time earlier than the oldest time.

  In S34, based on the device ID, LBA, and size included in the duplex request received by the second memory storage control unit 22 from the first processor 11, the data body to be duplexed included in the duplex request is the hard disk 15 of the first processor. It is checked whether or not the data of the disk area to be stored is stored in the second memory 24.

  In S35, as a result of the check in S34, the second memory storage control unit 22 stores in the second memory 24 the data of the disk area in which the data body to be duplexed included in the duplexing request is to be stored in the hard disk 15. It is determined whether or not. If stored (Yes), the process proceeds to S36, and if not stored (No), the process proceeds to S37.

  In S <b> 36, the second memory storage control unit 22 determines whether or not a new memory area need not be allocated when the data to be written in the write request is stored in the second memory 24. If it is not necessary to allocate a new memory area (Yes), the process proceeds to S37, and if a new memory area needs to be allocated (No), the process proceeds to S40.

  The reason for performing the processing of S34 to S36 is also the same as the processing of S6 to S8 described above, and the contents of the memory area allocation processing are the same as those described with reference to FIGS. .

  In S <b> 37, the second memory storage control unit 22 attempts to allocate a memory area for storing data included in the duplex request in the second memory 24.

  In S <b> 38, the second memory storage control unit 22 determines whether or not the second memory 24 has successfully allocated a memory area for storing data included in the duplex request. If the allocation is successful (Yes), the process proceeds to S40, and if the allocation fails (No), the process proceeds to S39.

  In S <b> 39, the second memory storage control unit 22 returns an NG response that has failed to duplex to the second memory 24 to the first processor 11.

  In S <b> 40, the second memory storage control unit 22 stores the device ID, LBA, size, and data body included in the duplex request in an appropriate memory area of the second memory 24. Note that “store in an appropriate memory area” here means that when a new memory area is allocated, it is stored in the new memory area as in S12 described above, and the allocation of a new memory area is unnecessary. If there is, it indicates that the data body included in the duplex request is overwritten on the data body of the existing data.

  In S41, the second memory storage control unit 22 overwrites the data included in the duplex request in the existing data memory area in S40, and overwrites the entire data body of the existing data in the second memory 24. When this happens, the storage time of the existing data is overwritten with the storage time included in the duplex request.

  In S <b> 42, the second memory storage control unit 22 returns an OK response indicating that duplication has been normally completed to the host computer 3.

[7. Hardware configuration]
FIG. 10 shows an example of a hardware configuration of an information processing apparatus in which the first computer 1 and the second computer 2 described above are embodied. The information processing apparatus includes a processor 901, a memory 902, a storage 903, a portable storage medium driving device 904, an input / output device 905, and a communication interface 906.

  The processor 901 includes a control unit, an arithmetic unit, an instruction decoder, and the like. The execution unit follows the instructions of the program decoded by the instruction decoder, and performs arithmetic / logic using the arithmetic unit according to a control signal output from the control unit. Perform the operation. The processor 901 has a TLB that functions as a control register that stores various types of information used for control, a cache that can temporarily store the contents of the memory 2 that has already been accessed, and a page table cache of virtual memory. Prepare. The processor 901 may have a configuration in which a plurality of CPU (Central Processing Unit) cores are provided. In the first computer 1, the first processor 11 corresponds to the processor 901, and in the second computer 2, the second processor 21 corresponds to the processor 901.

  The memory 902 is a storage device such as a RAM (Random Access Memory), for example. The memory 902 is a main memory in which a program executed by the processor 901 is loaded and data used for processing of the processor 901 is stored. In the first computer 1, the first memory 14 corresponds to the memory 902, and in the second computer 2, the second memory 24 corresponds to the memory 902.

  The storage 903 is a storage device such as an HDD (Hard Disk Drive) or a flash memory, and stores programs and various data. Note that the hard disk 15 of the first computer 1 corresponds to the storage 903.

  The portable storage medium driving device 904 is a device that reads data and programs stored in the portable storage medium 907. The portable storage medium 907 is, for example, a magnetic disk, an optical disk, a magneto-optical disk, or a flash memory. Note that the processor 901 executes a program stored in the storage 903 or the portable storage medium 907 in cooperation with the memory 902 or the storage 903. Note that the program executed by the processor 901 and data to be accessed may be stored in another device that can communicate with the information processing device.

The input / output device 905 is, for example, a keyboard or a display, and accepts an operation command by a user operation or the like while outputting a processing result by the information processing device.
The communication interface 906 is a LAN card, for example, and enables data communication with the outside. Each component of the information processing apparatus described above is connected by a bus 908.

[8. Effects and the like according to this embodiment]
According to the configuration and processing of the above-described embodiment, the first processor 11 transmits the oldest time that is the storage time of the oldest data stored in the first memory 14 to the second processor 21. For this reason, in the second processor 21, among the data stored in the second memory 24, data whose storage time is earlier than the oldest time is identified as a deletion target, and a plurality of data is deleted at once. be able to. In addition, even if there is only one data whose storage time is earlier than the oldest time (that is, only the data stored last in the hard disk 15), at least according to the embodiment described above. The first processor 11 does not need to transmit a data deletion request to the second processor 21 each time data in the first memory 14 is stored in the hard disk 15. For this reason, the transmission amount of this deletion request can be suppressed. Therefore, the processing load can be reduced as a whole system.

  In the above-described embodiment, the oldest time is added to the duplex request for data stored in the first memory 14 and transmitted to the second processor 21. Thereby, it is not necessary to transmit a deletion request for the purpose of deleting only the data in the second memory 24, and the amount of communication between the first processor and the second processor can be further suppressed.

  Further, in the above-described embodiment, the memory area allocation determination is performed in S6 to S8 and S34 to S36. That is, for example, when the first processor 11 stores the data included in the write request in the first memory 14, the data stored in the same disk area as the data already stored in the first memory 14 The data body of data is overwritten with the data body to be stored. As a result, allocation of a new memory area can be avoided and the number of times data is stored in the hard disk 15 can be suppressed. The same applies to the memory allocation process for the second memory 24.

  In the above-described embodiment, when the second processor 21 fails to allocate the memory area of the second memory 24, the first processor 11 stores the data in the first memory 14 in the hard disk 15, and the oldest time is set. After renewal, a duplex request including the oldest time is transmitted again to the second memory 24 to duplicate data. Thereby, duplication of data to the second memory 24 can be more reliably performed.

[9. Other Embodiments, Modifications, etc.]
Here, the first processor 11 is not limited to the oldest time, and any information may be transmitted as identification information as long as the data stored in the hard disk 15 at the end can be identified by the second processor 21. For example, the first processor 11 may add a sequence number indicating the order in which data is stored in the first memory 14 to the data, store the data in the first memory 14, and duplicate the data in the second memory. Then, the sequence number of the data having the oldest sequence number among the data stored in the first memory 14 may be transmitted to the second processor 21 instead of the oldest time.

  If no data is stored in the first memory 14, the first processor 11 transmits an identifier that can identify that the data is 0 instead of the oldest time, for example. May be. The second processor 21 may delete all data in the second memory 24 when the identifier is received.

  In the above-described embodiment, the oldest time is added to the duplex request for data stored in the first memory 14 and transmitted to the second processor 21. However, at the predetermined timing regardless of the duplex request. The oldest time may be transmitted to the second processor 21. Even in this way, the second processor 21 can delete a plurality of data in the second memory 24 collectively. Therefore, the data in the second memory 24 is more efficiently transmitted than when the deletion request is transmitted every time data is stored in the hard disk 15 in order from the oldest data in the first memory 14, and the data in the second memory 24 is deleted each time. Can be deleted.

  Furthermore, the present technology may naturally be applied to newly allocating the memory area of the first memory 14 to all the data without performing the memory area allocation determination in S6 to S8 and S34 to S36 described above. The purpose of can be achieved.

  The first processor 11 waits for a response from the second processor 21 and then stores data in the first memory 14. For example, the first processor 11 stores data in the first memory 14 first. Thereafter, a duplex request may be transmitted to the second processor 21. Further, the second processor 21 may perform the data deletion process of the second memory 24 before the data storage process of the second memory 24, or what is the data storage process of the second memory 24? It may be performed at an irrelevant timing.

  Further, in the above-described embodiment, as a case where data is stored in the first memory 14, only a case where a data write request from the host computer 3 is received is described. However, for example, the present technology can naturally be applied when a write request is generated from a program or the like executed by the first computer 1 itself.

  In the above-described embodiment, the first processor 11 and the second processor 21 are installed in different computers, but may be installed in one computer. In this case, data communication between the first processor 11 and the second processor 21 is performed via a bus or the like.

  In the above-described embodiment, the system that performs the write-back process using the first memory 14 and the hard disk 15 has been described as an example. However, the storage unit that can perform the storage process at high speed and the storage in comparison with this are described. Any device may be used as long as it has a relationship with a storage unit that takes time for processing. For example, the present technology may be applied to duplex processing in a system that performs write-back processing between a cache memory and a main memory. In addition, the first memory 14 and the second memory 24 may be installed outside the first processor 11 and the second processor 21 in structure.

  Further, for example, the hard disk 15 of the first computer 1 may be included in another device different from the first computer 1. In this case, the first computer 1 and the other device are communicably connected to the network, and the first computer 1 accesses the hard disk 15 via the network.

  The functional configuration and physical configuration of the information processing apparatus described above are not limited to the above-described aspects. For example, the functions and physical resources are integrated and mounted, or conversely, they are further distributed and mounted. It is also possible.

DESCRIPTION OF SYMBOLS 1 ... 1st computer, 2 ... 2nd computer, 3 ... Host computer, 11 ... 1st processor, 12 ... 1st memory storage control part, 13 ... Disk storage control part, 14 ... 1st memory, 15 ... Hard disk, 21 ... 2nd processor, 22 ... 2nd memory storage control part, 23 ... Deletion part, 24 ... 2nd memory

Claims (8)

In an information processing system including a first node and a second node,
The first node is:
Means for sequentially storing data in the first storage means, and transmitting data to be stored in the first storage means to the second node;
Means for sequentially reading out data stored in the first storage means from the first storage means and storing the data in the second storage means;
With
The second node is
Means for storing the data to be stored transmitted from the first node in a third storage means;
An identification that allows the first node to identify specific data stored in the second storage means lastly when notifying the second node about data that is allowed to be deleted Sending information to the second node;
The second node deletes the specific data identified by the information from the data stored in the third storage unit and the data stored before the specific data.
An information processing system characterized by this. The first node is:
When data is sequentially stored in the first storage means, storage order information is stored in association with each other,
When transmitting the data to be stored, the storage order information associated with the data to be stored is added and transmitted,
The identification information is storage order information associated with the specific data,
The second node is
The specific data is identified based on the storage order information associated with the specific data, and the one or more data whose storage order is earlier than the specific data is identified, and the deletion is performed. I do,
The information processing system according to claim 1. The first node performs transmission of the identification information in synchronization with transmission of the data to be stored.
The information processing system according to claim 1 or 2. The storage order information is a storage time when data is stored in the first storage means.
The information processing system according to claim 2. The first node stores the storage target when other data stored in the area of the second storage unit in which the data to be stored is stored is already stored in the first storage unit. Overwrite the other data with the data of
The second node may store the storage target when other data stored in the area of the second storage unit in which the storage target data is stored is already stored in the third storage unit. Overwrite the other data with the data of
The information processing system according to any one of claims 1 to 4, wherein When the second node stores the data to be stored in the third storage means and fails to allocate the area of the third storage means, the second node performs storage processing on the first node. Reply with a response indicating failure,
When the first node receives the response from the second node, the first node reads the data stored in the first storage unit from the first storage unit, stores the data in the second storage unit, and Sending the identification information to the second node again,
The information processing system according to any one of claims 1 to 5, wherein In an information processing system including a first node and a second node,
In the first node,
The data is sequentially stored in the first storage means, and data to be stored in the first storage means is transmitted to the second node,
Sequentially reading data stored in the first storage means from the first storage means and storing the data in the second storage means;
Let the process run,
In the second node,
Storing the storage target data transmitted from the first node in a third storage means;
Let the process run,
An identification that allows the first node to identify specific data stored in the second storage means lastly when notifying the second node about data that is allowed to be deleted Sending information to the second node;
The second node deletes the specific data identified by the information from the data stored in the third storage unit and the data stored before the specific data.
An information processing program characterized by that. In an information processing system including a first node and a second node,
The first node is
The data is sequentially stored in the first storage means, and data to be stored in the first storage means is transmitted to the second node,
Sequentially reading data stored in the first storage means from the first storage means and storing the data in the second storage means;
Execute the process,
The second node is
Storing the storage target data transmitted from the first node in a third storage means;
Execute the process,
An identification that allows the first node to identify specific data stored in the second storage means lastly when notifying the second node about data that is allowed to be deleted Sending information to the second node;
The second node deletes the specific data identified by the information from the data stored in the third storage unit and the data stored before the specific data.
An information processing method characterized by the above.